1. Field of the Invention
The present invention relates to a solid-state imaging element typified by a CCD and a CMOS image sensor, and a camera system.
2. Description of the Related Art
It is desired to realize a wide dynamic range solid-state imaging element by which even details can be so imaged that high-luminance information does not collapse and low-luminance part of the subject image does not become too black even in photographing against light such as light of car headlights, illumination light of a ball game ground, or sunlight.
Under such circumstances, regarding a solid-state imaging element such as a CCD, techniques for enlarging the dynamic range are disclosed in e.g. Japanese Patent No. 2125710, Japanese Patent Laid-Open No. Hei 3-117281, Japanese Patent Laid-Open No. Hei 9-205589, and Japanese Patent Laid-Open No. 2004-320119 (Hereinafter, Patent Document 1, Patent Document 2, Patent Document 3, and Patent Document 4, respectively).
Patent Document 1 discloses an example to enlarge the dynamic range of a CCD or the like by disposing plural regions (cells) different from each other in the sensitivity characteristic in one pixel of the CCD and allowing the pixel to have a so-called knee characteristic, in which the input-output characteristic changes in a stepwise manner.
The knee characteristic refers to a characteristic curve obtained by setting the slope of the high-input region gentler than that of the low-input region in a curve indicating the relationship between the exposure amount and the output current. The knee characteristic is often employed as a high-luminance signal compression technique.
As a method for changing the sensitivity of the sensitive region (cell), e.g. a method of changing the aperture ratio of the element, a method of providing an optical filter (ND filter), and a method of changing the impurity concentration are disclosed.
According to Patent Document 1, it seems that this technique can be applied also to XY address type imaging elements other than the CCD, although detailed description is not absent therein.
Patent Document 2 discloses an example to achieve such a wide dynamic range as to prevent the collapse of a subject image even under highlighting light such as bulb light. In this example, adjacent pixels or cells different from each other in the photosensitivity characteristic in a photosensitive pixel cell of a CCD are combined as one group, and the signal charges of the respective cells in one pixel are added to each other to thereby obtain the signal charge of the pixel.
In this case, as a scheme for changing the photosensitivity, cells having different pixel areas are combined as one group for example.
In the technique disclosed in Patent Document 3, one pixel of a photosensitive pixel cell in a CCD is divided into two regions different from each other in the sensitivity similarly, and the signal charges of the regions having the different sensitivities in the same pixel are mixed in a vertical register and are vertically transferred. Furthermore, in this technique, the signal charges of the regions having the different sensitivities are sorted into two horizontal transfer gates by a sorting gate. In addition, the signal of the higher sensitivity side is clipped by an external signal processing circuit and then is added to the signal of the lower sensitivity side to thereby form a video signal.
In this case, the characteristic graph of the video signal output with respect to the incident light amount is a polygonal line graph. In this graph, the slope on the higher sensitivity side (lower illuminance side) is steep and the slope on the lower sensitivity side (higher illuminance side) is gradual.
Patent Document 4 discloses a method for addressing a problem that, in an imaging element including a high-sensitivity imaging cell and a low-sensitivity imaging cell, the amount of the data of a RAW image (raw data) is large due to data by both the imaging cells.
Specifically, information on a captured image is analyzed and whether or not the image information of high-luminance part needs to be recorded is automatically determined. If it is determined that the image information of high-luminance part needs to be recorded, the RAW image data of the high-luminance part as well as the information of low-luminance part is recorded. If it is determined that the image information of high-luminance part does not need to be recorded, the information of the high-luminance part is not recorded but only the RAW image data of the low-luminance part is recorded.
One pixel is obtained by combining a main photosensitive pixel cell (it has large area and high sensitivity, and the center part of a microlens is mainly used as it) and a sub photosensitive pixel cell (it has small area and low sensitivity and is disposed on the edge side of the microlens) with each other.
Japanese Patent Laid-Open No. 2005-278135 (Patent Document 5) discloses a CMOS image sensor in which each of column parallel ADCs is composed of a comparator and an up/down counter. This CMOS image sensor can execute addition operation of pixel digital values for plural rows without additional circuits such as an adder and a line memory device.